Iir Tapped Cascade Lattice Structure

Tunable IIR Digital Filters
• We have described earlier two 1st-order and

two 2nd-order IIR digital transfer functions
with tunable frequency response
characteristics
• We shall show now that these transfer
functions can be realized easily using
allpass structures providing independent
tuning of the filter parameters
Copyright © 2001, S. K. Mitra
Tunable Lowpass and
Highpass Digital Filters
• We have shown earlier that the 1st-order
lowpass transfer function
1     z 1 
H LP ( z )  1   1

2 1  z 
and the 1st-order highpass transfer function
1    1  z 1 
H HP ( z )   1

2 1   z 
are doubly-complementary pair
Tunable Lowpass and
• Moreover, they can be expressed as
H LP ( z )  1 [1  A1( z )]
2
H HP ( z )  1 [1  A1( z )]
2
where
   z 1
A1( z ) 
1   z 1
is a 1st-order allpass transfer function
Tunable Lowpass and
• A realization of H LP (zand
) H HP (zbased
) on
the allpass-based decomposition is shown
below
H HP (z )
• The 1st-order allpass filter can be realized

using any one of the 4 single-multiplier allpass
structures described earlier

Tunable Lowpass and
• One such realization is shown below in
which the 3-dB cutoff frequency of both
lowpass and highpass filters can be varied
simultaneously by changing the multiplier
coefficient 

Tunable Lowpass and
• Figure below shows the composite
magnitude responses of the two filters for
two different values of 
1
a = 0.4
0.8 a = 0.05
Magnitude
0.6
0.4
0.2
0
0 0.2 0.4 0.6 0.8 1
w /p
Tunable Bandpass and
Bandstop Digital Filters
• The 2nd-order bandpass transfer function
1   1  z 2 
H BP ( z )   1  2

2  1  (1  ) z   z 
and the 2nd-order bandstop transfer function
1   1   z 1  z  2 
H BS ( z )   1  2


2  1  (1  ) z  pair
also form a doubly-complementary z 

• Thus, they can be expressed in the form
H BP ( z )  1 [1  A2 ( z )]
2
H BS ( z )  1 [1  A2 ( z )]
2
where
  (1  ) z 1  z  2
A2( z ) 
1  (1  ) z 1   z  2
is a 2nd-order allpass transfer function
• A realization of H BP (zand
) H BS (zbased
) on
the allpass-based decomposition is shown
below
• The 2nd-order allpass filter is realized using a

cascaded single-multiplier lattice structure

• The final structure is as shown below
• In the above structure, the multiplier  controls

the center frequency and the multiplier 
controls the 3-dB bandwidth

• Figure below illustrates the parametric
tuning property of the overall structure
 = 0.5  = 0.8
1 1
a = 0.4 b = 0.8
0.8 a = 0.05 0.8 b = 0.1
Magnitude
Magnitude
0.6 0.6
0.4 0.4
0.2 0.2
0 0
0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1
w /p w /p
IIR Tapped Cascaded Lattice
Structures
Realization of an All-pole IIR Transfer
Function
• Consider the cascaded lattice structure
derived earlier for the realization of an
allpass transfer function

Structures
• A typical lattice two-pair here is as shown
below
Wm1 ( z ) Wm (z )
S m1 ( z ) S m (z )
• Its input-output relations are given by

Wi ( z )  Wm1( z )  km z 1Sm ( z )
Sm1( z )  km Wm ( z )  z 1Sm ( z )
Structures
• From the input-output relations we derive
the chain matrix description of the two-pair:
W ( z )  1 k z 1  W ( z )
 i 1   i  i 
 Si 1 ( z )  k z  1 
 
 S i ( z ) 

i
• The chain matrix description of the
cascaded lattice structure is therefore
X ( z )  1 k z 1   1 k z 1   1 k z 1  W ( z )
 1  3 2 1  1 
 Y1 ( z )  k3 z 1  k2 z 1  k1 z 1   S1 ( z ) 
Structures
• From the above equation we arrive at
X1 ( z )  {1  [k1 (1  k2 )  k2k3 ]z 1
 [k2  k1k2 (1  k2 )]z  2  k3 z  2}W1( z )
 (1  d1z 1  d 2 z 2  d3 z 3 )W1( z )
using the relation S1 ( z )  W1 ( z ) and the
relations
k1  d1" , k2  d 2' , k3  d3

Structures
• The transfer function W1( z ) / X1( z ) is thus an
all-pole function with the same denominator
as that of the 3rd-order allpass function A3( z ):
W1 ( z ) 1

X1( z ) 1  d

Structures
Gray-Markel Method
• A two-step method to realize an Mth-order
arbitrary IIR transfer function
H ( z )  PM ( z ) / DM ( z )
• Step 1: An intermediate allpass transfer
function A ( z )  z  M D ( z 1 ) / D ( z )is
M M M
realized in the form of a cascaded lattice
structure
Structures
• Step 2: A set of independent variables are
summed with appropriate weights to yield
the desired numerator PM (z )
• To illustrate the method, consider the
realization of a 3rd-order transfer function
P3 ( z ) p0  p1z 1  p2 z  2  p3 z 3
H ( z)  
D3 ( z ) 1  d1z 1  d 2 z  2  d3 z 3

Structures
• In the first step, we form a 3rd-order allpass
transfer function
A3( z )  Y1( z ) / X1( z )  z 3 D3 ( z 1 ) / D3 ( z )
• Realization of A3(z ) has been illustrated
earlier resulting in the structure shown below

Structures
• Objective: Sum the independent signal
variables Y1 , S1 , S2 , and S3 with weights { i }
as shown below to realize the desired
numerator P3(z )

Structures
• To this end, we first analyze the cascaded
lattice structure realizing and determine the
transfer functions S1( z ) / X1( z ) , S2 ( z ) / X1( z ) ,
and S3 ( z ) / X1( z )
X1
Y1
• We have already shown
S1 ( z ) 1

X1 ( z ) D 3( z )
Structures
• From the figure it follows that
S2 ( z )  (k1  z 1)S1( z )  (d1"  z 1)S1( z )
and hence
S2 ( z ) d1"  z 1

X1( z ) D3( z )

Structures
• In a similar manner it can be shown that
S 3( z )  (d 2'  d1' z 1  z 2 )S1( z )
• Thus,
1 2
S 3( z ) d 2' '
 d1z  z

X 1( z ) D3 ( z )
• Note: The numerator of is precisely
the numerator of the allpassStransfer
i ( z ) / X1function
( z)
Ai( z )  Si ( z ) / Wi ( z )
Structures
• We now form
Yo ( z ) Y1 ( z ) S3 ( z ) S2 ( z ) S1 ( z )
 1  2  3  4
X1 ( z ) X1 ( z ) X1 ( z ) X1 ( z ) X1 ( z )

Structures
• Substituting the expressions for the various
transfer functions in the above equation we
arrive at
1 2 3
1(d3  d 2 z  d1z  z )
' ' 1 2 " 1
Yo ( z )   2 (d 2  d1z  z )  3 (d1  z )   4

X 1( z ) D3 ( z )

Structures
• Comparing the numerator of Yo ( z ) / X1( z )
with the desired numerator P3 (z ) and
1
equating like powers of z we obtain
1d3   2d 2'   3d1"   4  p0
1d 2   2d1'  3  p1
1d1   2  p2
1  p3

Structures
• Solving the above equations we arrive at
1  p3
 2  p2  1d1
 3  p1  1d 2   2d1'
 4  p0  1d3   2d 2'   3d1"

Structures
• Example - Consider
P3 ( z ) 0.44 z 1  0.362 z 2  0.02 z 3
H ( z)  
D3 ( z ) 1  0.4 z 1  0.18 z  2  0.2 z 3
• The corresponding intermediate allpass
transfer function is given by
z 3 D3 ( z 1 )  0.2  0.18 z 1  0.0.4 z  2  z 3
A3( z )  
D3 ( z ) 1  0.4 z 1  0.18 z  2  0.2 z 3

Structures
• The allpass transfer function A3( z ) was
realized earlier in the cascaded lattice form
as shown below
X1
Y1
• In the figure,
k3  d3  0.2, k2  d 2'  0.2708333
k1  d1"  0.3573771
Structures
• Other pertinent coefficients are:
d1  0.4, d 2  0.18, d3   0.2, d1'  0.4541667
p0  0, p1  0.44, p2  0.36, p3  0.02,
• Substituting these coefficients in
1  p3
 2  p2  1d1
 3  p1  1d 2   2d1'
 4  p0  1d3   2d 2'   3d1"
Structures
1  0.02,  2  0.352
3  0.2765333,  4   0.19016
• The final realization is as shown below
k1  0.3573771, k2  0.2708333, k3   0.2

Tapped Cascaded Lattice
Realization Using MATLAB
• Both the pole-zero and the all-pole IIR
cascaded lattice structures can be developed
from their prescribed transfer functions
using the M-file tf2latc
• To this end, Program 6_4 can be employed

Tapped Cascaded Lattice
• The M-file latc2tf implements the

reverse process and can be used to verify
the structure developed using tf2latc
• To this end, Program 6_5 can be employed

FIR Cascaded Lattice
Structures
• An arbitrary Nth-order FIR transfer function
of the form
N
H N ( z )  1  n 1 pn z  n
can be realized as a cascaded lattice structure
as shown below

Structures
• From figure, it follows that
X m ( z )  Xm1( z )  km z 1Ym1 ( z )
Ym ( z )  km Xm1 ( z )  z 1Ym 1( z )
• In matrix form the above equations can be written
as
 X m ( z )   1 km z 1   X m 1 ( z )
 Ym ( z )  k z 1   Ym 1 ( z ) 
where m
m  1, 2, ..., N
Structures
• Denote
Xm ( z ) Ym ( z )
Hm ( z )  , Gm ( z ) 
X 0 ( z) X 0 ( z)
• Then it follows from the input-output
relations of the m-th two-pair that
Hm ( z )  Hm 1 ( z )  km z 1Gm 1 ( z )
Gm ( z )  km Hm1( z )  z 1Gm1 ( z )
Structures
• From the previous equation we observe
H1 ( z )  1  k1z 1, G1 ( z )  k1  z 1
where we have used the facts
H0 ( z )  X 0 ( z ) / X 0 ( z )  1
G0 ( z )  Y0 ( z ) / X 0 ( z )  X 0 ( z ) / X 0 ( z )  1
• It follows from the above that
• z 1mirror-image
G1 ( z )is the 1H ( z 1 )
( z k1  1)  zof 1
G1 ( z ) H1 ( z )
Structures
• From the input-output relations of the m-th
two-pair we obtain for m = 2:
H2 ( z )  H1 ( z )  k2 z 1G1 ( z )
G2 ( z )  k2 H1( z )  z 1G1 ( z )
• Since H1 ( z ) and G1 ( z ) are 1st-order
polynomials, it follows from the above that
H 2 ( z ) and G2 ( z ) are 2nd-order polynomials

Structures
• Substituting G1 ( z )  z 1H1 ( z in1 )the two
previous equations we get
H2 ( z )  H1( z )  k2 z 2 H1( z 1)
G2 ( z )  k2 H1 ( z )  z  2 H1 ( z 1 )
• Now we can write
G2 ( z )  k2 H1 ( z )  z  2 H1 ( z 1 )
 z  2[k2 z 2 H1 ( z )  H1 ( z 1 )]  z  2 H 2 ( z 1 )
• is the mirror-image of
G2 ( z ) H 2 ( z)
Structures
• In the general case, from the input-output
relations of the m-th two-pair we obtain
Hm ( z )  Hm 1 ( z )  km z 1Gm 1 ( z )
Gm ( z )  km Hm1 ( z )  z 1Gm1 ( z )
• It can be easily shown by induction that
Gm ( z )  z m Hm ( z 1), m  1, 2, ..., N  1, N
• Gm (z ) is the mirror-image of Hm (z )

Structures
• To develop the synthesis algorithm, we
express H m 1( z ) and Gm1 ( z ) in terms of
H m(z ) and G m (z ) for m  N , N  1, ...,2,1
arriving at
1
HN 1 ( z )  2 { HN ( z )  k N GN ( z )}
(1 k N )
1
GN 1 ( z )  2 1 { k N HN ( z )  GN ( z )}
(1 k N ) z

Structures
• Substituting the expressions for
N
H N ( z )  1  n 1 pn z  n
and
N 1
GN ( z )  z H N ( z )  n 0 pn z  n  z  N
 N 1
in the first equation we get

1 N 1 n
H N 1 ( z )  2
{(1  k p
N N )   n 1 ( pn  k p
n N n ) z
1  kN
 ( pN  k N ) z  N }
Structures
• If we choose k N  p N, then HN 1 ( reduces
z)
to an FIR transfer function of order and can
N written
be 1 in the form
N 1
HN 1 ( z )  1  n 1 pn' z  n
where pn  k N p N  n
pn 
'
2 , 1  n  N 1
1 k N
• Continuing the above recursion algorithm, all
multiplier coefficients of the cascaded lattice
structure can be computed
Structures
• Example - Consider
H4 ( z )  1  1.2 z 1  1.12 z 2  0.12 z 3  0.08 z  4
• From the above, we observe k4  p4   0.08
• Using
pn  k 4 p4  n
pn 
'
2 , 1 n  3
1 k4
we determine the coefficients of H3 ( z ) :
p3'  0.2173913, p2'  1.2173913
p1'  1.2173913
Structures
• As a result,
H3 ( z )  1  1.2173913 z 1  1.2173913 z  2
 0.2173913 z 3
• Thus, k3  p3'  0.2173913
• Using
pn'  k3 p2'  n
pn"  2 , 1 n  2
1 k3
we determine the coefficients of H2 ( z ):
p2"  1.0, p1"  1.0
Structures
• As a result, H2 ( z )  1  z 1  z 2
• From the above, we get k2  p2"  1
• The final recursion yields the last multiplier
coefficient k1  p1" /(1  k2 )  0.5
• The complete realization is shown below
k1  0.5, k2  1, k3  0.2173913, k4   0.08

• The M-file tf2latc can be used to
compute the multiplier coefficients of the
FIR cascaded lattice structure
• To this end Program 6_6 can be employed
• The multiplier coefficients can also be
determined using the M-file poly2rc


Iir Tapped Cascade Lattice Structure

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Iir Tapped Cascade Lattice Structure

Uploaded by

Copyright:

Available Formats

Tunable IIR Digital Filters

• We have described earlier two 1st-order and

• The 1st-order allpass filter can be realized

Copyright © 2001, S. K. Mitra

Copyright © 2001, S. K. Mitra

Copyright © 2001, S. K. Mitra

• The 2nd-order allpass filter is realized using a

Copyright © 2001, S. K. Mitra

• In the above structure, the multiplier  controls

Copyright © 2001, S. K. Mitra

Copyright © 2001, S. K. Mitra

• Its input-output relations are given by

Copyright © 2001, S. K. Mitra

Copyright © 2001, S. K. Mitra

Copyright © 2001, S. K. Mitra

Copyright © 2001, S. K. Mitra

Copyright © 2001, S. K. Mitra

Copyright © 2001, S. K. Mitra

Copyright © 2001, S. K. Mitra

Copyright © 2001, S. K. Mitra

Copyright © 2001, S. K. Mitra

Copyright © 2001, S. K. Mitra

Copyright © 2001, S. K. Mitra

k1  0.3573771, k2  0.2708333, k3   0.2

Copyright © 2001, S. K. Mitra

• The M-file latc2tf implements the

Copyright © 2001, S. K. Mitra

Copyright © 2001, S. K. Mitra

Copyright © 2001, S. K. Mitra

Copyright © 2001, S. K. Mitra

Copyright © 2001, S. K. Mitra

in the first equation we get

k1  0.5, k2  1, k3  0.2173913, k4   0.08

Copyright © 2001, S. K. Mitra

You might also like